The genetics of early-onset Alzheimer's disease implicate the amyloid precursor protein (APP) and its derivatives in pathogenesis of the disease, yet we have not identified the mechanism through which familial APP variants lead to neuronal dysfunction and cognitive decline. Even fundamental mechanistic insight such as whether mutant APP primarily damages the cell in which it's expressed or neighboring cells through secreted fragments has been elusive. Cultured neurons and organotypic slices have been used to explore this issue, but it remains unclear whether observations made in immature neurons tested in vitro extend to mature neurons in the adult brain. Additionally, culture studies are limited in duration and may not reflect the consequences of chronic exposure present in the human disease. Longer exposures can be tested in vivo, however most transgenic models currently available were designed to provide ubiquitous APP over-expression throughout the brain, precluding distinction of cell-autonomous and cell-extrinsic effects. My laboratory is developing a novel mouse model that will provide the spatial resolution needed to separate cell-intrinsic and - extrinsic effects of mutant APP as well as the temporal resolution to distinguish the consequences of acute and chronic exposure. Our approach is based on a combination of viral and standard transgenesis to create mosaic animals in which the expression of mutant APP can be controlled through a virally-expressed tetracycline-transactivator. By adjusting the viral titer, we can control the degree of transgenic mosaicism in the brain to separate cell-intrinsic and cell-extrinsic effects of APP over-expression. Incorporation of the tetracycline transactivator provides rapid and reversible temporal control over the onset and duration of transgenic APP over-expression. We provide preliminary data showing the development of the new model and describe the next steps in its optimization. We then propose experiments using the new mosaic tet-off APP mice to address a critical question that has been un-testable in existing models: Does APP over- expression alter neuronal structure and function in a cell-autonomous or a cell-extrinsic manner? We will examine the impact of APP over-expression on the structure and function of transgenic neurons and their wild- type neighbors. The new model will allow us to move this question from the culture dish into the mammalian brain where it can be examined with greater temporal freedom and better fidelity to the human disease. PUBLIC HEALTH RELEVANCE: This study will create a new animal model for Alzheimer's disease that will help us understand how inherited mutations in familial forms of the disease cause neuronal dysfunction thought to underlie dementia. The new model will be used to test whether the Alzheimer's-related amyloid precursor protein damages the cells in which it's made or acts on nearby neighbors through secreted fragments, akin to asking whether neurons in the Alzheimer's brain die by suicide or murder. A better understanding of this fundamental issue will improve our ability to design and deliver effective therapeutics for the disease.